Import everything you need so as to setup the environment.
In [1]:
import matplotlib.image as mpimg
import numpy as np
import cv2
from skimage.feature import hog
import matplotlib.pyplot as plt
import glob
import time
import pickle
import os
from moviepy.editor import VideoFileClip
from IPython.display import HTML
from scipy.ndimage.measurements import label
from sklearn.svm import LinearSVC
from sklearn.preprocessing import StandardScaler
from skimage.feature import hog
from sklearn.model_selection import train_test_split
%matplotlib inline
Using a bunch of functions from the lessons. These are the functions used to extract features out of the test images
In [2]:
def color_transform(img, color_space):
if color_space != 'RGB':
if color_space == 'HSV':
output_img = cv2.cvtColor(img, cv2.COLOR_RGB2HSV)
elif color_space == 'LUV':
output_img = cv2.cvtColor(img, cv2.COLOR_RGB2LUV)
elif color_space == 'HLS':
output_img = cv2.cvtColor(img, cv2.COLOR_RGB2HLS)
elif color_space == 'YUV':
output_img = cv2.cvtColor(img, cv2.COLOR_RGB2YUV)
elif color_space == 'YCrCb':
output_img = cv2.cvtColor(img, cv2.COLOR_RGB2YCrCb)
elif color_space == 'RGB2YCrCb':
output_img = cv2.cvtColor(img, cv2.COLOR_RGB2YCrCb)
elif color_space == 'BGR2YCrCb':
output_img = cv2.cvtColor(img, cv2.COLOR_BGR2YCrCb)
elif color_space == 'RGB2LUV':
output_img = cv2.cvtColor(img, cv2.COLOR_RGB2LUV)
else: output_img = np.copy(img)
return output_img
def bin_spatial(img, size=(32, 32)):
color1 = cv2.resize(img[:,:,0], size).ravel()
color2 = cv2.resize(img[:,:,1], size).ravel()
color3 = cv2.resize(img[:,:,2], size).ravel()
return np.hstack((color1, color2, color3))
def color_hist(img, nbins=32):
# Compute the histogram of the RGB channels separately
rhist = np.histogram(img[:,:,0], bins=nbins)
ghist = np.histogram(img[:,:,1], bins=nbins)
bhist = np.histogram(img[:,:,2], bins=nbins)
# Generating bin centers
bin_edges = rhist[1]
bin_centers = (bin_edges[1:] + bin_edges[0:len(bin_edges)-1])/2
# Concatenate the histograms into a single feature vector
hist_features = np.concatenate((rhist[0], ghist[0], bhist[0]))
# Return the individual histograms, bin_centers and feature vector
return hist_features, rhist, ghist, bhist, bin_centers
def get_hog_features(img, orient, pix_per_cell, cell_per_block,
vis=False, feature_vec=True):
# Call with two outputs if vis==True
if vis == True:
features, hog_image = hog(img, orientations=orient, pixels_per_cell=(pix_per_cell, pix_per_cell),
cells_per_block=(cell_per_block, cell_per_block), transform_sqrt=True,
visualise=vis, feature_vector=feature_vec)
return features, hog_image
# Otherwise call with one output
else:
features = hog(img, orientations=orient, pixels_per_cell=(pix_per_cell, pix_per_cell),
cells_per_block=(cell_per_block, cell_per_block), transform_sqrt=True,
visualise=vis, feature_vector=feature_vec)
return features
# Extract features for an image; Note input is an image file
def extract_features_single(img, color_space='RGB', spatial_size=(32, 32),
hist_bins=32, orient=9,
pix_per_cell=8, cell_per_block=2, hog_channel=0,
spatial_feat=True, hist_feat=True, hog_feat=True, vis=False):
#1) Define an empty list to receive features
img_features = []
#2) Apply color conversion if other than 'RGB'
feature_image = color_transform(img, color_space)
#3) Compute spatial features if flag is set
if spatial_feat == True:
spatial_features = bin_spatial(feature_image, size=spatial_size)
#4) Append features to list
img_features.append(spatial_features)
#5) Compute histogram features if flag is set
if hist_feat == True:
hist_features, _, _, _, _ = color_hist(feature_image, nbins=hist_bins)
#6) Append features to list
img_features.append(hist_features)
#7) Compute HOG features if flag is set
if hog_feat == True:
if hog_channel == 'ALL':
hog_features = []
for channel in range(feature_image.shape[2]):
hog_features.append(get_hog_features(feature_image[:,:,channel],
orient, pix_per_cell, cell_per_block,
vis=False, feature_vec=True))
hog_features = np.concatenate(hog_features)
else:
if vis==True:
hog_features, hog_image = get_hog_features(feature_image[:,:,hog_channel], orient,
pix_per_cell, cell_per_block, vis=True, feature_vec=True)
else:
hog_features = get_hog_features(feature_image[:,:,hog_channel], orient,
pix_per_cell, cell_per_block, vis=False, feature_vec=True)
#8) Append features to list
img_features.append(hog_features)
#9) Return concatenated array of features
if vis==True:
return np.concatenate(img_features), hog_image
else:
return np.concatenate(img_features)
#Extract features for a bunch of images five a file path
def extract_features(imgs, color_space='RGB', spatial_size=(32, 32),
hist_bins=32, orient=9,
pix_per_cell=8, cell_per_block=2, hog_channel=0,
spatial_feat=True, hist_feat=True, hog_feat=True):
# Create a list to append feature vectors to
features = []
# Iterate through the list of images
for file in imgs:
image = mpimg.imread(file)
file_features = extract_features_single(image, color_space, spatial_size, hist_bins, orient, pix_per_cell,
cell_per_block, hog_channel, spatial_feat, hist_feat, hog_feat)
features.append(file_features)
# Return list of feature vectors
return features
"""
#Handy function to analyze images; Found this on a stackoverflow
# page that I cant find now
def plot_test_image(fig, rows, cols, imgs, titles):
for i, img in enumerate(imgs):
plt.subplot(rows, cols, i+1)
plt.title(i+1)
img_dims = len(img.shape)
if img_dims < 3:
plt.imshow(img, cmap='hot')
plt.title(titles[i])
else:
plt.imshow(img)
plt.title(titles[i])
"""
# Draw boxes on an image given the bboxes co-ordinates
def draw_boxes(img, bboxes, color=(0, 0, 255), thick=6):
# Make a copy of the image
draw_img = np.copy(img)
# Iterate through the bounding boxes
for bbox in bboxes:
# Draw a rectangle given bbox coordinates
cv2.rectangle(draw_img, bbox[0], bbox[1], color, thick)
# Return the image copy with boxes drawn
return draw_img
In [3]:
# Read in cars and notcars from the udacity data set provided. This can take very long
# and hence I restrict the number of images. Accuracy seems to vary
# depending on the images used.
cars = []
noncars = []
# populate cars
image_types = os.listdir("vehicles/")
for imtype in image_types:
cars.extend(glob.glob("vehicles/" + imtype + '/*'))
# populate non-cars
image_types = os.listdir("non-vehicles/")
for imtype in image_types:
noncars.extend(glob.glob("non-vehicles/" + imtype + '/*'))
print('Number of cars:', len(cars))
print('Number of non-cars:', len(noncars))
#Make sure there are images in array
fig = plt.figure()
fig.add_subplot(1,2,1)
plt.imshow(mpimg.imread(cars[0]))
fig.add_subplot(1,2,2)
plt.imshow(mpimg.imread(noncars[0]));
In [4]:
#Plot some car images
fig=plt.figure(figsize=(16,16))
for i in range(0,64):
ax = fig.add_subplot(8,8,i+1)
idx = np.random.randint(len(cars));
plt.imshow(mpimg.imread(cars[idx]));
plt.axis('off')
In [5]:
#Plot some non-car images
fig=plt.figure(figsize=(16,16))
for i in range(0,64):
ax = fig.add_subplot(8,8,i+1)
idx = np.random.randint(len(noncars));
plt.imshow(mpimg.imread(noncars[idx]));
plt.axis('off')
In [6]:
### TODO: Tweak these parameters and see how the results change.
color_space = 'RGB' # Can be RGB, HSV, LUV, HLS, YUV, YCrCb
orient = 6 # HOG orientations
pix_per_cell = 8 # HOG pixels per cell
cell_per_block = 2 # HOG cells per block
hog_channel = 0 # Can be 0, 1, 2, or "ALL"
spatial_size = (16, 16) # Spatial binning dimensions
hist_bins = 16 # Number of histogram bins
spatial_feat = True # Spatial features on or off
hist_feat = True # Histogram features on or off
hog_feat = True # HOG features on or off
car_idx = np.random.randint(0, len(cars))
non_car_idx = np.random.randint(0, len(noncars))
car_img = mpimg.imread(cars[car_idx])
non_car_img = mpimg.imread(noncars[non_car_idx])
f, car_hog_image = extract_features_single(car_img, color_space=color_space, spatial_size=spatial_size,
hist_bins=hist_bins, orient=orient,
pix_per_cell=pix_per_cell,
cell_per_block=cell_per_block, hog_channel=hog_channel,
spatial_feat=spatial_feat,
hist_feat=hist_feat, hog_feat=hog_feat, vis=True)
f, non_car_hog_image = extract_features_single(non_car_img, color_space=color_space,
spatial_size=spatial_size,
hist_bins=hist_bins, orient=orient,
pix_per_cell=pix_per_cell,
cell_per_block=cell_per_block,
hog_channel=hog_channel,
spatial_feat=spatial_feat,
hist_feat=hist_feat, hog_feat=hog_feat, vis=True)
fig = plt.figure(figsize=(12,3))
plt.subplot(1, 4, 1)
plt.imshow(car_img)
plt.subplot(1, 4, 2)
plt.imshow(car_hog_image, cmap='hot')
plt.subplot(1, 4, 3)
plt.imshow(non_car_img)
plt.subplot(1, 4, 4)
plt.imshow(non_car_hog_image, cmap='hot');
Final set of parameters after experimenting with several different values to improve the accuracy
In [7]:
color_space = 'YCrCb' # Can be RGB, HSV, LUV, HLS, YUV, 'YCrCb'
orient = 9
pix_per_cell = 8
cell_per_block = 2
hog_channel = 'ALL' # Can be 0, 1, 2, 'ALL'
spatial_size = (16, 16) # down sample size
hist_bins = 32
spatial_feat = True
hist_feat = True
hog_feat = True
car_features = extract_features(cars, color_space=color_space, spatial_size=spatial_size,
hist_bins=hist_bins, orient=orient,
pix_per_cell=pix_per_cell, cell_per_block=cell_per_block, hog_channel=hog_channel,
spatial_feat=spatial_feat, hist_feat=hist_feat, hog_feat=hog_feat)
non_car_features = extract_features(noncars, color_space=color_space, spatial_size=spatial_size,
hist_bins=hist_bins, orient=orient,
pix_per_cell=pix_per_cell, cell_per_block=cell_per_block, hog_channel=hog_channel,
spatial_feat=spatial_feat, hist_feat=hist_feat, hog_feat=hog_feat)
X = np.vstack((car_features, non_car_features)).astype(np.float64)
# Fit a per-column scaler
X_scaler = StandardScaler().fit(X)
# Apply the scaler to X
scaled_X = X_scaler.transform(X)
# Define the labels vector
y = np.hstack((np.ones(len(car_features)), np.zeros(len(non_car_features))))
# Split up data into randomized training and test sets
rand_state = np.random.randint(0, 100)
X_train, X_test, y_train, y_test = train_test_split(
scaled_X, y, test_size=0.2, random_state=rand_state)
print('Using:',orient,'orientations',pix_per_cell,
'pixels per cell and', cell_per_block,'cells per block')
print('Feature vector length:', len(X_train[0]))
# Use a linear SVC
svc = LinearSVC()
# Check the training time for the SVC
%time svc.fit(X_train, y_train)
# Check the score of the SVC
print('Test Accuracy of SVC = ', round(svc.score(X_test, y_test), 4))
In [8]:
# Define a function that takes an image, start and stop positions in both x and y,
# window size (x and y dimensions), and overlap fraction (for both x and y)
def slide_window(img, x_start_stop=[None, None], y_start_stop=[None, None],
xy_window=(64, 64), xy_overlap=(0.5, 0.5)):
# If x and/or y start/stop positions not defined, set to image size
if x_start_stop[0] == None:
x_start_stop[0] = 0
if x_start_stop[1] == None:
x_start_stop[1] = img.shape[1]
if y_start_stop[0] == None:
y_start_stop[0] = 0
if y_start_stop[1] == None:
y_start_stop[1] = img.shape[0]
# Compute the span of the region to be searched
xspan = x_start_stop[1] - x_start_stop[0]
yspan = y_start_stop[1] - y_start_stop[0]
# Compute the number of pixels per step in x/y
nx_pix_per_step = np.int(xy_window[0]*(1 - xy_overlap[0]))
ny_pix_per_step = np.int(xy_window[1]*(1 - xy_overlap[1]))
# Compute the number of windows in x/y
nx_buffer = np.int(xy_window[0]*(xy_overlap[0]))
ny_buffer = np.int(xy_window[1]*(xy_overlap[1]))
nx_windows = np.int((xspan-nx_buffer)/nx_pix_per_step)
ny_windows = np.int((yspan-ny_buffer)/ny_pix_per_step)
# Initialize a list to append window positions to
window_list = []
# Loop through finding x and y window positions
# Note: you could vectorize this step, but in practice
# you'll be considering windows one by one with your
# classifier, so looping makes sense
for ys in range(ny_windows):
for xs in range(nx_windows):
# Calculate window position
startx = xs*nx_pix_per_step + x_start_stop[0]
endx = startx + xy_window[0]
starty = ys*ny_pix_per_step + y_start_stop[0]
endy = starty + xy_window[1]
# Append window position to list
window_list.append(((startx, starty), (endx, endy)))
# Return the list of windows
return window_list
In [9]:
# Define a function you will pass an image
# and the list of windows to be searched (output of slide_windows())
def search_windows(img, windows, clf, scaler, color_space='RGB',
spatial_size=(32, 32), hist_bins=32,
hist_range=(0, 256), orient=9,
pix_per_cell=8, cell_per_block=2,
hog_channel=0, spatial_feat=True,
hist_feat=True, hog_feat=True):
#1) Create an empty list to receive positive detection windows
on_windows = []
#2) Iterate over all windows in the list
for window in windows:
#3) Extract the test window from original image
test_img = cv2.resize(img[window[0][1]:window[1][1], window[0][0]:window[1][0]], (64, 64))
#4) Extract features for that window using single_img_features()
features = extract_features_single(test_img, color_space=color_space,
spatial_size=spatial_size, hist_bins=hist_bins,
orient=orient, pix_per_cell=pix_per_cell,
cell_per_block=cell_per_block,
hog_channel=hog_channel, spatial_feat=spatial_feat,
hist_feat=hist_feat, hog_feat=hog_feat)
#5) Scale extracted features to be fed to classifier
test_features = scaler.transform(np.array(features).reshape(1, -1))
#6) Predict using your classifier
prediction = clf.predict(test_features)
#7) If positive (prediction == 1) then save the window
if prediction == 1:
on_windows.append(window)
#8) Return windows for positive detections
return on_windows
In [10]:
# Process an image taking in an image, a y range, a window size,
# and the overlaps of the sliding window
def process_image(img, y_start_stop=[None, None], xy_window=(128, 128), overlap=.5):
draw_img = np.copy(img)
img = img.astype(np.float32)/255
print(np.min(img), np.max(img))
windows = slide_window(img, x_start_stop=[None, None], y_start_stop=y_start_stop,
xy_window=xy_window, xy_overlap=(overlap, overlap))
t1 = time.time()
hot_windows = search_windows(img, windows, svc, X_scaler, color_space=color_space,
spatial_size=spatial_size, hist_bins=hist_bins,
orient=orient, pix_per_cell=pix_per_cell, cell_per_block=cell_per_block,
hog_channel=hog_channel, spatial_feat=spatial_feat,
hist_feat=hist_feat, hog_feat=hog_feat)
print(time.time()-t1, 'seconds to process a large image with', len(windows), 'windows')
window_img = draw_boxes(draw_img, hot_windows, color=(0,0,255), thick=6)
return window_img
In [12]:
# attempt to draw windows on 6 test_images
searchpath = 'test_images/*'
test_images = glob.glob(searchpath)
y_start_stop=[None, None]
xy_window=(128, 128)
overlap=.5
fig = plt.figure(figsize=(12,18))
idx=1
for filename in test_images:
img = mpimg.imread(filename)
window_img = process_image(img, y_start_stop, xy_window, overlap)
plt.subplot(5, 2, idx)
plt.imshow(window_img)
plt.title(filename)
idx+=1
These values seem to detect most cars. Unable to detact the car in test3 no matter how much I have tuned the parameters
In [13]:
searchpath = 'test_images/*'
test_images = glob.glob(searchpath)
y_start_stop=[380,650]
xy_window=(128, 128)
overlap=.5
fig = plt.figure(figsize=(12,18))
idx=1
for filename in test_images:
img = mpimg.imread(filename)
window_img = process_image(img, y_start_stop, xy_window, overlap)
plt.subplot(5, 2, idx)
plt.imshow(window_img)
plt.title(filename)
idx+=1
In [14]:
from scipy.ndimage.measurements import label
def get_labels(heatmap):
return label(heatmap)
def draw_labeled_bboxes(img, labels):
# Iterate through all detected cars
for car_number in range(1, labels[1]+1):
# Find pixels with each car_number label value
nonzero = (labels[0] == car_number).nonzero()
# Identify x and y values of those pixels
nonzeroy = np.array(nonzero[0])
nonzerox = np.array(nonzero[1])
# Define a bounding box based on min/max x and y
bbox = ((np.min(nonzerox), np.min(nonzeroy)), (np.max(nonzerox), np.max(nonzeroy)))
# Draw the box on the image
cv2.rectangle(img, bbox[0], bbox[1], (0,0,255), 6)
# Return the image
return img
# a function that adds "heat" to a map for a list of bounding boxes
def generate_heat(heatmap, bbox_list):
# Iterate through list of bboxes
for box in bbox_list:
# Add += 1 for all pixels inside each bbox
# Assuming each "box" takes the form ((x1, y1), (x2, y2))
heatmap[box[0][1]:box[1][1], box[0][0]:box[1][0]] += 1
def find_cars(img, y_start_stops, x_start_stops, scales, cells_per_steps=[2], pix_per_cell=8, orient=9):
img_boxes = []
heatmap = np.zeros_like(img[:,:,0])
# normalize
img = img.astype(np.float32)/255
window_count = 0
t = time.time()
for y_start_stop, x_start_stop, scale, cells_per_step in zip(y_start_stops, x_start_stops, scales, cells_per_steps):
# crop to boundaries passed in
cropped_img = img[y_start_stop[0]:y_start_stop[1],x_start_stop[0]:x_start_stop[1],:]
# convert to new color
cropped_img = color_transform(cropped_img, color_space='RGB2YCrCb')
if scale != 1:
imshape = cropped_img.shape
resized_img = cv2.resize(cropped_img, (np.int(imshape[1]/scale), np.int(imshape[0]/scale)))
else:
resized_img = cropped_img
ch1 = resized_img[:,:,0]
ch2 = resized_img[:,:,1]
ch3 = resized_img[:,:,2]
hog1 = get_hog_features(ch1, orient, pix_per_cell, cell_per_block, feature_vec=False)
hog2 = get_hog_features(ch2, orient, pix_per_cell, cell_per_block, feature_vec=False)
hog3 = get_hog_features(ch3, orient, pix_per_cell, cell_per_block, feature_vec=False)
nxblocks = (ch1.shape[1] // pix_per_cell) - 1
nyblocks = (ch1.shape[0] // pix_per_cell) - 1
nfeat_per_block = orient*cell_per_block**2
window=64
nblocks_per_window = (window // pix_per_cell) - 1
# cells_per_step defines overlap. 1 - cells_per_step / pix_per_cell gives us overlap percentage
nxsteps = (nxblocks - nblocks_per_window) // cells_per_step
nysteps = (nyblocks - nblocks_per_window) // cells_per_step
for xb in range(nxsteps):
for yb in range(nysteps):
window_count += 1
ypos = yb*cells_per_step
xpos = xb*cells_per_step
# extract HOG for patch
hog_feat1 = hog1[ypos:ypos+nblocks_per_window, xpos:xpos+nblocks_per_window].ravel()
hog_feat2 = hog2[ypos:ypos+nblocks_per_window, xpos:xpos+nblocks_per_window].ravel()
hog_feat3 = hog3[ypos:ypos+nblocks_per_window, xpos:xpos+nblocks_per_window].ravel()
hog_features = np.hstack((hog_feat1, hog_feat2, hog_feat3))
xleft = xpos*pix_per_cell
ytop = ypos*pix_per_cell
# extract the image patch for spatial and hist
subimg = cv2.resize(resized_img[ytop:ytop+window, xleft:xleft+window], (64,64))
# use spatial and color hist features
spatial_features = bin_spatial(subimg, size=spatial_size)
hist_features, _, _, _, _ = color_hist(subimg, nbins=hist_bins)
# stack all 3 features
feature_stack = np.hstack((spatial_features, hist_features, hog_features))
test_features = X_scaler.transform(feature_stack.reshape(1, -1))
# prediction with SVM
pred = svc.predict(test_features)
if pred == 1:
xbox_left = np.int(xleft*scale) + x_start_stop[0]
ytop_draw = np.int(ytop*scale)
win_draw = np.int(window*scale)
top_left = (xbox_left, ytop_draw+y_start_stop[0])
bottom_right = (xbox_left+win_draw, ytop_draw+win_draw+y_start_stop[0])
img_boxes.append((top_left, bottom_right))
generate_heat(heatmap, img_boxes)
return heatmap
def pipeline(image):
out_img, windows = find_cars(image, ystart, ystop, scale, svc, X_scaler, orient,
pix_per_cell, cell_per_block, spatial_size, hist_bins)
heat = np.zeros_like(image[:,:,0]).astype(np.float)
# Add heat to each box in box list
heat = add_heat(heat,windows)
# Apply threshold to help remove false positives
heat = apply_threshold(heat,2)
# Visualize the heatmap when displaying
heatmap = np.clip(heat, 0, 255)
# Find final boxes from heatmap using label function
labels = label(heatmap)
draw_img = draw_labeled_bboxes(np.copy(image), labels)
return draw_img
In [15]:
# Apply threshold to help remove false positives
def apply_threshold(heatmap, threshold):
# Zero out pixels below the threshold
heatmap[heatmap <= threshold] = 0
# Return thresholded map
return heatmap
def process_single_image(img):
# parameters
y_start_stops = [(400,475),(400,500),(400,500),(400, 650),(400,650)]
x_start_stops = [(325, 950),(0,550),(700, 1280),(950, 1280),(0, 350)]
scales = [.75, 1.25, 1.25, 2, 2]
cells_per_steps = [2, 1, 1, 1, 1]
heatmap = find_cars(img, y_start_stops, x_start_stops, scales, cells_per_steps)
# threshold to minimize false positives
apply_threshold(heatmap, threshold=3)
# handy function as mentioned to generate labels from contiguous regions in heatmap
labels = label(heatmap)
draw_img = draw_labeled_bboxes(np.copy(img), labels)
return draw_img
In [16]:
fig = plt.figure(figsize=(12,18))
idx=1
for img_src in glob.glob('test_images/*'):
img = mpimg.imread(img_src)
processed_img = process_single_image(img)
plt.subplot(5,2,idx)
plt.imshow(processed_img)
plt.title(img_src)
idx+=1
In [17]:
last_n = 5
heatmaps = [None] * last_n
iteration = 0
from functools import reduce
def pipeline(img):
global iteration, last_n, heatmaps
# parameters
y_start_stops = [(400,475),(400,500),(400,500),(400, 650),(400,650)]
x_start_stops = [(325, 950),(0,550),(700, 1280),(950, 1280),(0, 350)]
scales = [.75, 1.25, 1.25, 2, 2]
cells_per_steps = [2, 1, 1, 1, 1]
heatmap = find_cars(img, y_start_stops, x_start_stops, scales, cells_per_steps)
heatmaps[iteration % last_n] = heatmap
if iteration >= last_n:
integrated_heatmap = reduce(lambda x, y: x+y, heatmaps)
else:
integrated_heatmap = heatmap
# threshold to minimize false positives
apply_threshold(integrated_heatmap, threshold=25)
# handy function as mentioned to generate labels from contiguous regions in heatmap
labels = label(heatmap)
draw_img = draw_labeled_bboxes(np.copy(img), labels)
iteration += 1
return draw_img
In [18]:
last_n = 5
heatmaps = [None] * last_n
iteration = 0
output_v1 = 'test_video_output.mp4'
clip1 = VideoFileClip("test_video.mp4")
test_clip = clip1.fl_image(pipeline)
%time test_clip.write_videofile(output_v1, audio=False)
HTML("""
<video width="800" height="600" controls>
<source src="{0}">
</video>
""".format(output_v1))
Out[18]:
In [19]:
last_n = 5
heatmaps = [None] * last_n
iteration = 0
output_v1 = 'project_video_output.mp4'
clip1 = VideoFileClip("project_video.mp4")
test_clip = clip1.fl_image(pipeline)
%time test_clip.write_videofile(output_v1, audio=False)
HTML("""
<video width="800" height="600" controls>
<source src="{0}">
</video>
""".format(output_v1))
Out[19]:
In [ ]: